Smart Sensors for Condition Based Maintenance: a Test Case in the Manufacturing Industry

Condition Based Maintenance (CBM) is a well-known concept and it has been demonstrated that it is the way ahead to prognostic maintenance for failure avoidance and for the reduction of maintenance cost. This paper presents an application for Condition Based Maintenance, with a specific focus on State Detection, according to MIMOSA OSA-CBM reference architecture. The papers aims at presenting peculiarity of development of such a kind of solution when considering the use of Smart Sensors instead of traditional devices. Indeed, breakthrough in CBM is expected from the development of ICT and embedded systems. This technology supply integrated chips implementing all the necessary circuitry to manage field data capture, data processing, local diagnosis, local feedback (where possible) and information transfer to the upper control levels. These so-called smart sensors exploit new technologies of micro sensors (MEMS, micro electro mechanical systems) and wireless communication together with the computing power of a microprocessor. In particular, applications related to maintenance and human safety appear to be very promising due to the unstructured nature of these domains, where self-configuring networks of intelligent devices can better comply with an ever changing and partially unpredictable environment. A test case is deployed on a typical manufacturing equipment: a robot. The objective of the test case presented by the paper is not to develop new diagnostic algorithms, but to implement some statistical analysis within a monitoring infrastructure built with Smart Sensors. The case of analysis that the paper will present grounds on the use of wireless sensor devices for temperature measures gathered on the electric motors of the robot. Then, data are transmitted through a wireless network to a receiver unit that accomplishes also elaboration by using statistical methods and then, thanks to a web-service communication, results are made available to external requests and users. An advisory is generated when something is out of the normal behaviour of the equipment. Finally, the user can check this information through the Human Machine Interface available via web-service.